Combined Inhibition of Notch Signaling and Bcl-2/Bcl-xL Results in Synergistic Antimyeloma Effect

Multiple myeloma (MM) is characterized by the uncontrolled proliferation of malignant plasma cells in bone marrow. Despite significant improvements in the treatment of this disease with novel therapeutic agents including bortezomib and lenalidomide, many patients ultimately become resistant to therapy. Therefore, searching for new molecular targets and drug combinations is critical for the development of effective therapeutic options.

Signaling through the receptor/transcriptional regulator Notch plays an important role in the survival of tumor cells and their protection from apoptosis induced by chemotherapeutic agents. The Notch pathway has been shown to be activated in the majority of solid and hematological tumors studied, including MM (1–3). According to the conventional model, Notch signaling is initiated upon binding of the Notch receptors expressed by MM cells to the Notch ligands, Jagged or Delta, that are expressed by both MM cells and the surrounding cellular components of bone marrow microenvironment. This binding initiates 2 subsequent proteolytic cleavages within the transmembrane domain resulting in the liberation and translocation of the intracellular domain of Notch (ICN) into the nucleus followed by its binding to the transcriptional repressor CSL (CBF-1). Binding of ICN displaces corepressor complexes and recruits coactivators, thereby, turning CBF-1 into a transcriptional activator (4, 5). A second proteolytic cleavage that leads to Notch activation is regulated by a protease complex possessing γ-secretase activity. Pharmacological compounds are able to inhibit γ-secretase [γ-secretase inhibitors, γ-secretase inhibitors (GSI)], and therefore, Notch signaling have shown impressive preclinical activity and are currently being tested in clinical trials (6, 7). Our previous data showed that inhibition of Notch signaling with GSI resulted in significant cytotoxicity of MM cells. This effect was mediated through a dramatic upregulation of the proapoptotic bcl-2 family protein Noxa (8).

Noxa belongs to the BH3-only “sensitizer” protein family that act by displacing the BH3-only “activators” like Bid and Bim from the antiapoptotic proteins, allowing the activators to bind Bax and Bak (9). Alternatively, antiapoptotic proteins could directly inhibit Bax and Bak activation (10). It is well known that Noxa has a very high affinity for the antiapoptotic protein Mcl-1, but not Bcl-2, Bcl-xL, or Bcl-w. While activation of Bax is controlled by Bcl-2, Mcl-1, and Bcl-xL, Mcl-1 also cooperates with Bcl-xL to sequester Bak and prevent its activation (11, 12). Recent studies have indicated that targeting of both Mcl-1 and Bcl-xL is necessary in order to release Bak (13).

ABT-737, a small molecule BH3-only mimetic developed by Abbott Laboratories, specifically targets the antiapoptotic bcl-2 members Bcl-2, Bcl-xL and Bcl-w (14). This compound prevents sequestration of Bax and Bak by Bcl-2 and Bcl-xL and induces apoptosis of tumor cells including multiple myeloma cells (14–17). However, ABT-737 does not have a strong affinity for Mcl-1 and therefore, does not effectively trigger apoptosis of tumor cells expressing a high level of Mcl-1 (13, 18, 19). Recently, several groups have reported that the downregulation of Mcl-1 sensitized tumor cells to apoptosis induced by ABT-737 (18, 20–22). Binding of Noxa to Mcl-1 reduces Mcl-1 levels by promoting proteosomal degradation. Therefore, strategies to increase Noxa could be potentially advantageous for overcoming Mcl-1 related resistance to ABT-737.

Using MM cell lines and primary cells as well as an in vivo xenograft and a SCID-hu model of MM, we demonstrate that the combination of these 2 agents has a significant synergistic antimyeloma effect.

Human MM NCI-H929, U266, and RPMI-8226 cell lines were obtained from the American Type Culture Collection and were kept in culture no longer than 6 weeks. MM1S cell line was a gift from Dr. Steven Rosen (Northwestern University); no authentication was done by the authors. Cells were cultured as described previously (2). GSI (γ-secretase inhibitor XII) was purchased from Calbiochem and dissolved in DMSO (dimethyl sulfoxide) as a 5 mmol/L stock solution. Bcl-2/Bcl-xL inhibitor ABT-737 was provided by Abbott Laboratories and dissolved in DMSO as a 20 mmol/L stock solution for in vitro studies and in 30% propylene glycol, 5% Tween 80, and 65% D5W (5% dextrose in water) for in vivo studies. Chemical structures of GSI XII and ABT-737 are shown in Fig. 1. Pan-caspase inhibitor z-VAD-FMK was obtained from Bachem (Bachem Americas) and dissolved in DMSO.

Collection of BM (bone marrow) was in accordance with University of South Florida Institutional Review Board approved protocol. CD138⁺ myeloma cells were isolated from BM samples using CD138 MicroBeads and the MidiMACS magnetic cell separator according to the manufacturer's protocol (Miltenyi Biotec). Cells were cultured in α-minimal essential medium supplemented with 10% FBS with or without GSI, ABT-737, or their combination for 48 hours. Blood samples were obtained from healthy volunteers, and mononuclear cells were isolated by Ficoll-Paque gradient centrifugation and then cultured in α-minimal essential medium supplemented with 10% FBS.

Cells were treated with various concentrations of GSI and/or ABT-737 for 48 hours. 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) dye was added at the last 4 hours of incubation. Cells were then pelleted and the medium aspirated. Insoluble formazan complexes were solubilized with DMSO and the absorbance was measured at 540 nm using a Benchmark Plus microplate spectrophotometer (Bio-Rad). MM cell survival and the CI were calculated using CalcuSyn software (Biosoft). Each experimental condition was done in triplicate and done at least twice.

Apoptosis of myeloma cells was detected by the Annexin V binding assay as described earlier (2). Ten thousand events were acquired and analyzed using a FACSCalibur or LSR II flow cytometers (BD) and FlowJo software (Tree Star).

Activation of Bak and Bax were determined by flow cytometry as described previously using specific antibodies that recognize only conformationally active forms of these proteins [anti-Bax (clone 3, BD Biosciences) or anti-Bak (Ab-1, Calbiochem); ref. 23].

In order to detect cytochrome c release, cells were collected and permeabilized with 100 μL of digitonin (50 μg/mL) for 5 minutes on ice. Cells were then fixed with 4% paraformaldehyde followed by blocking (3% bovine serum albumin in PBS with 0.05% saponin) for 1 hour. Cells were incubated with an anti-cytochrome c antibody (BD Bioscience) overnight at 4°C followed by labeling with a secondary FITC-conjugated anti-mouse IgG antibody. Ten thousand events were acquired using a FACSCalibur flow cytometer (BD).

Western blotting was done as described previously (2, 24). The antibodies against the following proteins were used: Mcl-1, β-actin (Santa Cruz); Bcl-xL (BD Bioscience); Bcl-2, PARP, and cleaved caspase 3 (Cell Signaling Technology); and Noxa (clone 114C307, Calbiochem).

Control nontargeting pool siRNA or specific Noxa, Bak or Bax siRNA were obtained from Ambion. Myeloma RPMI 8226 cells were transfected with siRNA using Amaxa nucleofection technology (Lonza). Briefly, 5 × 10⁶ cells were resuspended in buffer V and electroporated using program A20. Cells were cultured overnight and then treated with GSI, ABT-737, or combination thereof.

Tumor tissues were fixed in formalin, embedded in paraffin, and then the slides were prepared. Immunohistochemical staining was done using Ventana Discovery XT automated system (Ventana Medical Systems). Heat-induced antigen retrieval method was used for cleaved caspase 3 staining and enzymatic retrieval method [4 minutes in Protease 1(Ventana)] was used for Noxa staining. Slides were then incubated with primary antibodies against cleaved caspase 3 (#9661, Cell Signaling, 1:400 dilution) or Noxa (#3665–100, Biovision, 1:25 dilution) for 1 hour followed by incubation with secondary anti-rabbit antibody (Ventana) for 20 minutes. Ventana OmniMap kit was used as a detection system. Slides were then counterstained with Hematoxylin.

SCID-beige (C.B-Igh-1b/GbmsTac-Prkdc^scid-Lyst^bg N7) mice were purchased from Taconic and kept in pathogen-free conditions in the animal facility of the H. Lee Moffitt Cancer Center. All experimental procedures were done in accordance with the guidelines of the University of South Florida Institutional Animal Care and Use Committee.

Xenograft model was established by subcutaneous inoculation of 1 × 10⁷ MM 8226 cells in the right flank of 6–8 week-old SCID‐beige mice. In approximately 3 weeks when the tumor became measurable, mice were assigned to 1 of the following 4 groups: control group, GSI or ABT-737 treated groups, or a group treated with the combination of GSI and ABT-737. GSI and ABT-737 were administered i.p. daily for 14 days at a dose of 5 mg/kg/day and 75 mg/kg/day, respectively. Tumor size was measured twice a week during and after treatment. Mice were followed up for approximately 2 weeks after finishing treatment.

SCID-hu model was established as described previously (8). Human fetal tissues from 20 to 23 weeks of gestation were obtained from Advance Bioscience Resource. Fragments of human fetal bones (humerus, femur, or tibia) were implanted s.c. in 6 week–old-female SCID-beige mice. Six weeks after implantation, 5 × 10⁴ 8226 myeloma cells in 50 μL PBS were injected directly into each implanted bone. Treatment with GSI, ABT-737, or the combination began approximately 4 weeks following tumor inoculation. GSI was administered at a dose of 5 mg/kg/day and ABT-737 at a dose of 100 mg/kg/day i.p. daily for 14 days. Mice in the control group did not receive injections. Blood (50–100 μL) was drawn from a submandibular vein and the level of free lambda light chain of human immunoglobulin was measured by an enzyme-linked immunosorbent assay (Bethyl) before and after treatment.

Statistical significance of differences observed in cells treated with single drug compared to cells treated with drug combination was determined using the Student t test. The P values less than 0.05 were considered significant. In order to assess for a possible synergistic effect of combined therapies CalcuSyn software was used. A combination index (CI) less than 0.9 indicates synergism, while CI between 0.9–1.0 indicates an additive effect (25). Two-way ANOVA and Mann-Whitney nonparametric test were used for statistical analysis of the differences between in vivo treatment groups and was calculated using GraphPad Prism software.

To investigate the combined effect of the Notch inhibitor GSI and ABT-737 on MM cells, we initially measured cellular viability using the MTT assay in different myeloma cell lines. Treatment with either GSI or ABT-737 alone resulted in decreased survival of myeloma cells, whereas the combination of these 2 compounds had a significantly stronger antimyeloma effect (Fig. 2A–C). The effect of the drug combination was evaluated by calculating the CI and considered to be synergistic when the CI value was less than 0.9. The combination of GSI and ABT-737 had a significant synergistic cytotoxic effect on all MM cell lines studied. The lowest CI values were 0.745 for 8226 cells, 0.63 for U266 and 0.35 for MM1S cells. The effect of the GSI combination with ABT-737 on primary CD138⁺ MM cells was similar to its effect on MM cell lines with CI between 0.4 and 0.7 (Fig. 2D).

Our previous data have demonstrated that treatment with GSI resulted in apoptosis of MM cells (8). ABT-737 was also shown to induce MM cell death (17). Therefore, to confirm the effect of the drug combination we next evaluated MM cell apoptosis using the Annexin V binding assay. A significant (P < 0.05) increase in apoptosis of cells treated with combination of GSI and ABT-737 was observed as compared to cells treated with either GSI or ABT-737 alone (Fig. 3A). To confirm the validity of the data obtained with cell lines we used primary CD138⁺ MM cells isolated from the bone marrow of 4 different patients with MM. Addition of ABT-737 increased the proapoptotic effect of GSI in primary CD138⁺ cells (Fig. 3B). In contrast, the cytotoxic effect of the GSI combination with ABT-737 was not observed in peripheral blood mononuclear cells isolated from healthy donors (Fig. 3C). Thus, taken together these results demonstrate that ABT-737 synergizes with GSI in induction of apoptosis in MM cells and does not affect normal mononuclear cells.

Initially we investigated whether the inhibition of Notch signaling with GSI affected the levels of the antiapoptotic members of the bcl-2 family. MM cells were treated with different concentrations of GSI and the levels of Bcl-2, Bcl-xL and Mcl-1 were evaluated by Western blotting. GSI only slightly downregulated the level of Bcl-xL and did not affect the level of Bcl-2 (Fig. 4A). Consistent with our previous data, inhibition of Notch resulted in dramatic upregulation of the proapoptotic BH3-only protein Noxa (Fig. 4A; ref 8). All MM cells studied expressed Mcl-1 protein and there was no substantial difference in Mcl-1 level between cell lines. We did not observe GSI-induced Mcl-1 degradation in MM cells (Fig. 4A).

To clarify the role of Noxa in apoptosis induced by drug combination, we downregulated its level in MM cells using specific siRNA (Fig. 4B). Cells were then treated with GSI, ABT-737, or combination of these compounds and apoptosis was detected by Annexin V/DAPI staining by flow cytometry. Consistent with our previous data (8), inhibition of Noxa abrogated the effect of GSI (Fig. 4B). Decrease in Noxa level also resulted in reduced ABT-737 cytotoxicity and significantly reduced cell death induced by combination of GSI with ABT-737 (Fig. 4B).

Inhibition of Bcl-2 and Bcl-xL with ABT-737 has been shown to induce caspase-dependent apoptosis (14, 21). Here we investigated whether ABT-737 synergizes with GSI in Bak and Bax activation followed by release of cytochrome c and cleavage of caspase 3 and PARP. Substantially increased proapoptotic conformational changes of Bak and Bax were observed in MM cells treated with the combination of GSI and ABT-737 as compared to cells treated with either GSI or ABT-737 alone (Fig. 4C). The combination of these 2 drugs was able to activate Bax in 1.8 ± 0.53-fold (range from 1.3- to 2.5-fold) and Bak in 1.9 ± 0.6-fold (range 1.4 to 2.8 fold) in MM cells when compared to the level of activation in cells treated with either GSI or ABT-737 alone. To clarify the role of Bak and Bax in apoptosis induced by GSI in combination with ABT-737, we downregulated expression of these proteins in the cells using siRNA. MM 8226 cells were transfected with either control nontargeting pool siRNA or specific Bak and Bax siRNA. This resulted in a substantial decrease of Bak and/or Bax protein levels (Fig. 4D). Cells were then treated with GSI, ABT-737 or drug combination and apoptosis was evaluated by Annexin V/DAPI staining. Depletion of Bak/Bax level significantly decreased cell death induced by either GSI or ABT-737 alone or in combination (Fig. 4D).

When activated, Bak and Bax can permeabilize the outer mitochondrial membrane and release proapoptogenic factors including cytochrome c that is needed to activate the caspases and induce DNA fragmentation. To confirm the downstream effect of Bak and Bax activation, we evaluated the release of cytochrome c in MM cells using flow cytometry. Treatment of 3 myeloma cell lines with the combination of GSI and ABT-737 resulted in a dramatic release of cytochrome c as compared to cells exposed to either GSI or ABT-737 alone (Fig. 5A). Release of cytochrome c was followed by the cleavage of caspase 3 and PARP (Fig. 5B). Thus, taken together these data demonstrate that the combination of GSI with ABT-737 is synergistic in inducing mitochondria-dependent apoptosis. Pretreatment of MM cells with the pan-caspase inhibitor z-VAD-FMK abrogated the effect of the GSI/ABT-737 drug combination as was evaluated by measuring MM cell apoptosis using the Annexin V binding assay (Fig. 5C) and release of cytochrome c by flow cytometry (Fig. 5D).

To investigate whether the addition of ABT-737 would enhance the antitumor effect of GSI in vivo, we used 2 different MM mouse models, the xenograft and SCID-hu model. Initial experiments were done using the xenograft model, where MM 8226 cells were injected s.c. into a flank of SCID-beige mice. When the tumor mass became palpable, mice were split into 4 groups and treated with GSI, ABT-737, or a combination thereof for 14 days. No apparent signs of toxicity such as behavioral changes or weight loss were observed in treated animals. At selected doses the antitumor effect of GSI or ABT-737 alone was rather moderate. The combination of these 2 drugs had significantly stronger antitumor effect (Fig. 6A). After treatment stops, a rapid tumor growth was observed in GSI and ABT-737 groups while tumor growth in combination group was significantly delayed (Fig. 6A). Immunohistochemical analysis of tumor tissues using antibodies against cleaved caspase 3 and Noxa indicated that GSI and ABT-737 induced substantially stronger apoptosis of tumor cells when used in combination than when used individually (Fig. 6B). The disadvantage of the xenograft model is that MM cells grow as a solid tumor, which is artificial for MM. Therefore, to evaluate the effect of the drug combination in the context of human BM, we also used a SCID-hu model. In this model 8226 MM cells were injected directly into previously engrafted fragments of human bone, and tumor growth was monitored by measuring the level of a specific human paraprotein in mouse serum. Mice were treated with GSI, ABT-737, or a combination of these 2 drugs. Serum level of free human lambda chain was used to evaluate tumor growth. Administration of each of these drugs alone decreased the level of free human lambda chain in serum when compared to a control group (Fig. 6C and D). The combination of GSI and ABT-737 had significantly more potent antimyeloma effect than each of the drug alone (Fig. 6D).

In this study, we examined the effect of simultaneous inhibition of Notch signaling and the bcl-2 family antiapoptotic members Bcl-2/Bcl-xL on MM cell growth and survival. We found that the combination of the Notch inhibitor GSI and the Bcl-2/Bcl-xL inhibitor ABT-737 had a synergistic cytotoxic effect on MM cells in vitro and in vivo.

Antiapoptotic members of bcl-2 family including Bcl-2, Bcl-xL, and Mcl-1 are highly overexpressed by myeloma cells and are considered to be among the major contributing factors to the survival, growth, and chemoresistance of these cells (26–30). Therefore targeting these proteins represents an attractive therapeutic strategy. ABT-737, a potent small molecule BH3 mimetic was recently developed by Abbott (14). As a single agent it has been shown to induce apoptosis in various solid and hematopoietic tumor cells including multiple myeloma (16, 17). ABT-737 binds with high affinity to Bcl-2, Bcl-xL, and Bcl-w of prosurvival bcl-2 family members (14). This binding results in activation (oligomerization) of Bax and Bak that subsequently leads to apoptosis (31). In order to liberate Bak, both Bcl-xL and Mcl-1 need to be neutralized (11, 12). Although ABT-737 disrupts the interaction of Bcl-xL with Bak, this compound has a low affinity for Mcl-1. As a result, tumor cells overexpressing Mcl-1 are resistant to this agent, and neutralization of Mcl-1 has been shown to be effective in enhancing the antitumor effect of ABT-737 (18, 31). According to our results RPMI-8226 cells were significantly more sensitive to ABT-737 as compared to H929 and U266 cells while the level of Mcl-1 in these cell lines was similar. The higher sensitivity of RPMI-8226 cells to ABT-737 could be explained by the fact that these cells have substantially lower level of Bcl-2 and higher level of Noxa expression.

One of the potent regulators of Mcl-1 is the proapoptotic BH3-only protein Noxa (12, 32). When overexpressed, Noxa specifically and tightly binds to Mcl-1 targeting it to proteasomal degradation, and displaces Bak. While GSI was able to significantly induce Noxa, we did not observe Mcl-1 degradation in myeloma cells following GSI treatment. Recent study indicated that it is rather Mcl-1 functional inactivation than degradation and elimination that is critical to activation of Bak (33). Although Mcl-1 level was not affected in our study, GSI was able to induce apoptosis and synergize with ABT-737 in antimyeloma effect confirming these data. Why Mcl-1 was not undergoing proteasomal degradation after dramatic induction of Noxa by GSI is not clear. However, it is possible that beside inhibition of Notch, GSI could potentially have off-target effects. Recently it was suggested that selected gamma-secretase inhibitors could also block proteasome activity (34). Therefore, one could speculate that degradation of Mcl-1 is abrogated in myeloma cells due to off-target effect of GSI on proteasome activity while the cytotoxic effect of the compound is preserved due to induction of Noxa and neutralization of Mcl-1.

We have previously showed that by blocking Notch signaling with GSI, Noxa is dramatically upregulated and induces activation of Bak in MM cell lines and primary patient myeloma cells (8). These data suggest that the antitumor effect of Notch inhibition could be substantially improved by pharmacological targeting of Bcl-2 and Bcl-xL. Indeed, our results showed that simultaneous inhibition of Notch and Bcl-2/Bcl-xL by using GSI and ABT-737 resulted in significant induction of Bak and Bax activation, followed by cytochrome c release, cleavage of caspase 3, and cell apoptosis. This is the first report showing synergism between inhibitions of Notch signaling and Bcl-2/Bcl-xL in vitro and in vivo using xenograft and SCID-hu model. Consistent with previous observations GSI or ABT-737 alone moderately decreased myeloma tumor growth in mice. When treatment was stopped GSI group did not retain significant antitumor effect that we observed previously using the same concentrations of the compound (8). This is most likely attributed to the fact that different mouse strains were used for xenograft model in these experiments. Our data indicated that MM 8226 tumor grew substantially slower in SCID/NOD mice as compared to SCID-beige mice (Fig. 6A; ref. 8). Although GSI had antitumor effect in both models, due to kinetics of tumor growth, the effect of GSI was less prominent in SCID-beige mice. Although at selected doses GSI and ABT-737 alone has only moderate effect on MM growth, the combination of these 2 drugs has statistically significant antitumor effect that is retained after the end of the treatment.

Our finding of synergism between Notch and Bcl-2/Bcl-xL inhibitors is consistent with the studies where the induction of Noxa with topoisomerase I inhibitor CPT-11 in colorectal cancer cells and with proteasome inhibitors in melanoma and lymphoid malignancies, sensitized these cells to ABT-737 (35–37).

In summary, we show for the first time that the inhibition of Notch synergistically cooperates with ABT-737 to enhance apoptosis in myeloma cells. Our findings indicate that this therapeutic strategy could be potentially beneficial for the treatment of patients with multiple myeloma.

No potential conflicts of interest were disclosed.

National Institute of Health grant CA130923 and Multiple Myeloma Research Foundation Senior Research Award (Y. Nefedova). This work was supported in part by the Flow Cytometry Core Facility at the H. Lee Moffitt Cancer Center.